Hello !
I am simulating forging process in which plastic metal has contact with rigid mold. Could someone help comment on
the difference between node-to-surface and surface-to-surface schemes and especially their performance in accuracy and efficiency ?
While reading the manual, I am confused with the face-to-face penalty method. It’s said “the master and
slave mesh are put on top of each other, the common areas, which are polygons (sides of quadratic elements are approximated by piecewise linear lines), are
identified and triangulated.” I wonder how this “common area” is determined. It appears that the contacting area is a portion of master surface. If so, then what boundary is the contact area ? In the schematic diagram attached below, where the third dimension can be thought of perpendicular to the screen, does this area range from 3’ to 4’ which are projected boundary of slave surface on master surface , or from 2 to 5, which are intersecting lines between master and slave, or from 1’ to 6’, which are also projections of slave boundary ? It’ll be good if some literature on face-to-face penalty method could be referred. I find very little on this method.
If I remember correctly, in one chapter of the PhD thesis of Saskia Sitzmann you will find plenty of details of the penalty F2F in CCX. It is available in the internet.
it seems theses by Jaro Hokkanen (Aalto University, 2014) who adviser by Dr. Guido Dhondt are more relevant as primary refferences. as i know, theses by Saskia Sitzmann (Universitat Erlangen-Nurnberg, 2016) are discussed more about advanced contact methods (Mortar).
the thesis by Jaro Hokkanen is the relevant one for surface-to-surface penalty, although the method of identifying the potential contact area with the appropriate integration points is common to the Mortar method (Saska Sitzmann). The idea is that the complete slave area is projected on the master area in the direction of the normal on the master area. So the potential contact area is this part of the slave area, for which there is an opposite master surface, in your drawing 1 to 6. It is identified once at the beginning of each increment, and the appropriate integration points are identified. These may be 100 or more per slave face. In each iteration used to get convergence for the increment, the integration points which are really in contact (i.e. with penetration) are identified and in these integration points spring elements are created.
